Created attachment 119259 [details] Gzipped tarball with relevant files for this bug Trying to install Ubuntu Linux 15.10 on an older machine and Xorg fails with the dreaded "Unknown Card-Ids" error. I saw that someone else on the Debian lists had a similar problem with other machine and solved it himself by adding the missing card ids to via_id.c and recompiling the driver. I did it myself to and got it working. I'm not a developer though, and much less a driver developer - so, instead of the cute diffs, I'm attaching the logs, info about my setup and my own via_id.c which worked for this rig. I only added one line to the latest release I got through anongit. Notice that the attached via_id.c is how I got it working, but: 1) By reading the C file a bit more carefully, I realize I kept the extra "VIA_DEVICE_LCD}" there probably by mistake. This board has only got one standard monitor out (i.e.: no DVI, HDMI or extra "notebook" out). It's your call, though. ;-) 2) I've used the board's DMI info (also splashed on the boot screen) to include the entry in via_id.c. Feel free to alter it if needed (you're the maintainers, after all). Thank you for maintaining this driver.
I still got this PC with me, in case any further testing is needed. Thanks again!
Just to let you know, OpenChrome project has been without a maintainer for many months. This is why you did not get a response. I recently obtained a commit privilege, and have already made several commits. https://cgit.freedesktop.org/openchrome/xf86-video-openchrome/ I also wrote a detailed instructions on how to compile the device driver. https://lists.freedesktop.org/archives/openchrome-devel/2016-February/001753.html It may not be necessary since you already know how to compile the device driver, but just included the location of it for your reference. I can say that your bug appears similar to a newer bug that got reported recently. https://bugs.freedesktop.org/show_bug.cgi?id=94130 It will be nice if you can try the latest OpenChrome master branch code without your LCD related workaround, in order to see what happens. In general, I am inclined to reduce the usage of the predefined device table moving forward.
Created attachment 121786 [details] attachment-8793-0.html Hello, Kevin. Don't worry at all. You're doing it for free, and I've been keeping track of the truckload of changes you've been comitting to OpenChrome. I have to say I'm pretty glad you got the time, interest *and* technical skills to help so thoroughly with this project. Thank you for all you (and everyone else) have been doing for this hardware. Believe it or not, this video card is quite widespread around (because it's so cheap) and it's sad to see Linux having so many issues with it. There are less issues now, thanks to your work (and the other contributors' too). And maybe kernel integration in the foreseeable future. Pretty cool. :-) The bad news is that I don't have the hardware with me anymore, so I really can't test the changes. It's not with anyone I'm in touch with either (it's been donated), so I can't contact anyone to test it. The good news is that, as soon as I "committed" my changes (attached to this bug) and rebuilt the driver, it worked flawlessly and automagically. No quirky Xorg settings, nothing. So it seems that it was *really* just a matter of adding it to the device table. My hardware id is strictly the same as the one in the new bug you mentioned. I know testing it would be the perfect solution, but there's little theoretical difference between our cases - the solution for the new bug should "solve" my problem as well. I'm glad you're trying to make the driver less table-dependent. I can't really understand why it would *need* to match a device table in order to work, once it has already been detected as OpenChrome by the whole subsystem (I can understand its usefulness, but not it's absolute *necessity*). My best regards to you and the team! De: "bugzilla-daemon@freedesktop.org" <bugzilla-daemon@freedesktop.org> Para: tylderon@yahoo.com.br Enviadas: Terça-feira, 16 de Fevereiro de 2016 3:08 Assunto: [Bug 92711] Unknown Card-Ids (3371|1019|2125) Comment # 2 on bug 92711 from Kevin Brace Just to let you know, OpenChrome project has been without a maintainer for many months. This is why you did not get a response. I recently obtained a commit privilege, and have already made several commits. https://cgit.freedesktop.org/openchrome/xf86-video-openchrome/ I also wrote a detailed instructions on how to compile the device driver. https://lists.freedesktop.org/archives/openchrome-devel/2016-February/001753.html It may not be necessary since you already know how to compile the device driver, but just included the location of it for your reference. I can say that your bug appears similar to a newer bug that got reported recently. https://bugs.freedesktop.org/show_bug.cgi?id=94130 It will be nice if you can try the latest OpenChrome master branch code without your LCD related workaround, in order to see what happens. In general, I am inclined to reduce the usage of the predefined device table moving forward. You are receiving this mail because: - You reported the bug.
Kevin, is there any testing I can do for you here? Or is http://bugs.freedesktop.org/show_bug.cgi?id=94130#c4 sufficient? (I can't tell if there's an actionable item here).
(In reply to Jeffrey Walton from comment #4) > Kevin, is there any testing I can do for you here? Or is > http://bugs.freedesktop.org/show_bug.cgi?id=94130#c4 sufficient? > > (I can't tell if there's an actionable item here). And by the way... I found the "Unknown Card-Ids (3371|1019|2125)" message was still present in the latest driver. However, the crash is gone, and that's a big win.
Since this bug is almost identical to Bug 94130, and I recommended that bug report's status to be changed to "RESOLVED FIXED," I will change the status of this bug report to "RESOLVED FIXED" as well. The soon to be released OpenChrome Version 0.3.4 will fix the crash issue. As for the unknown ID reported in the log, it does not affect the operation of the device driver, so it will not be dealt with for now. I will like OpenChrome to move in the direction of reducing the use of a known device table moving forward, so this issue will be handled when the code is updated to reduce the dependency on the known device table.
(In reply to Jeffrey Walton from comment #4) > Kevin, is there any testing I can do for you here? Or is > http://bugs.freedesktop.org/show_bug.cgi?id=94130#c4 sufficient? > > (I can't tell if there's an actionable item here). Hi Jeffrey, I think you did enough testing, as far as I am concerned. The latest master branch code fixed your "show stopper" bug, so I think you can close your bug report.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.